Analogy-making is the most beautiful manifestation of the fundamental property of human’s thought. It shows how the incoming information maps to the already memorized old one, and they both change until fit consistently. What we see depends on what we already know and vice versa, what we know depends on what we have already seen. It is broadly accepted that ability to make deep analogies is one of the strongest predictors of the intelligence, the creativity, the ability to understand causal relationships. However, analogy-making, being one of the strongest instrument of human though, may have a dark side too. There are examples in the literature how wrong analogies may produce various fallacies of human thought. Widespread beliefs in world conspiracy, superstitions, etc. could be a product of wrong analogies. Different effects of the known phenomena of confirmation bias are also modeled by the sub-processes of analogy-making. This paper systematizes different types of fallacies and possibilities how those effects could be modeled within a cognitive architecture are discussed.